They then conducted animal studies, including one in which the altered cancer cells were injected into the stomach wall. Gastric tumors grew in both a control group and in the experimental group of animals. At 22 days, tumors growing in the experimental group, in cells that expressed low levels of HIF-1, were significantly smaller than those growing in the control group.

"We were able to markedly inhibit tumor growth, and we also found something interesting," says Ellis. "The number of blood vessels between the groups was unchanged, yet in experimental mice, the vessels were smaller, and without a lumen. They didn't appear to be able to remain open. They were not 'good' vessels."

The investigators evaluated the structure of these altered vessels and found they didn't contain many pericytes. Normal blood vessels have endothelial cells surrounded by pericytes, which provide structurally rigidity and also modulate endothelial cell survival.  "We found that the endothelial cells had little pericyte coverage in tumors with decreased HIF-1 activity, suggesting that HIF-1 not only regulates molecules that mediate endothelial cell function, but also pericyte function," Ellis says.

These results suggest that HIF-1 not only regulates VEGF expression in cancer cells, he says, "but also contributes to the formation of a complex proangiogenic microenvironment in tumors, affecting both the structure and function of blood vessels."

The study was supported by grants from the National Institutes of Health. Co-authors include, from M. D. Anderson's Department of Cancer Biology: Oliver Stoeltzing, M.D., Fan Fan, BS, Weinbiao Liu, M.D., and Anna Belcheva, M.D.; from M. D. Anderson's Department of Surgical Oncology: Marya McCarty, Ph.D., and Jane Wey, M.D. Another co-author, Gregg Semenza, M.D., Ph.D., at the Johns Hopkins University School of Medicine, is a leading expert on HIF-1 research.

mdanderson